lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Tatu Saloranta <t...@hypermall.net>
Subject Re: Indexing Tips and Hints
Date Tue, 25 Feb 2003 15:52:08 GMT
On Tuesday 25 February 2003 03:48, Andrzej Bialecki wrote:
> petite_abeille wrote:
...
> > Multivalent's phelps.io.BufferedRandomAccessFile... and I'm happy to
> > report that... I doesn't seems to make a shred of difference in my
> > case... but as always YMMV.
>
> This is strange, or at least counter-intuitive - if you buffer larger
> parts of data in RAM than the standard implementation does, it should
> definitely be faster... Let's wait and see what Terry comes up with.
>
> BTW. how large indexes did you use for testing? Also, it could be that
> the indexing process is bound by some other bottleneck, and buffering
> helps only when searching already existing index.

Perhaps it also depends on platform -- on Linux (for example), all smallish 
files are very likely to be kept in memory, if accessed often. This because 
all non-allocated RAM is used for disk buffers automatically.
There is still syscall overhead for reading, but compared to actual disk reads 
it will be much faster.

-+ Tatu +-



---------------------------------------------------------------------
To unsubscribe, e-mail: lucene-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: lucene-user-help@jakarta.apache.org


Mime
View raw message